システム・アプローチに欠けるもの : 経済学における反省(<特集>システム論を問いなおす)

概要

論文の詳細を見る
The systems approach or systems theory was introduced into and welcomed by the economics field in the 1960's and 1970's. It is considered to bring a holistic viewpoint into economics and advance the analysis of complex systems such as a national economy or the world economy. More than 30 years have passed since then, and the fever seems to have faded away without making a substantial contribution to economics. This paper questions why the systems approach has contributed so little to the development of economics since the 1970's. One reason lies in the theoretical structure of economics itself. Microeconomics had its systems theory in the form of equilibrium theory and had no need to introduce a new framework from outside of its boundaries. The systems approach could not provide an analytical substitute that was superior to the equilibrium theory. Another reason lies in the lacuna of the systems theory. It has no framework that explains how to deal with individual actions and cannot provide a guide analyzing complex phenomena that results from the interactions of many human agents. The systems approach could provide a map of institutions and functions but could not explain how they work when they are arranged to interact with each other. This inability explains why rational choice theory, in spite of its purely individualistic inclination, is now spreading from economics to sociology and political science. The systems approach should incorporate a behavioral theory and reorganize its framework in such a way that individual choice and behavior regain its proper place in the functional analysis of the whole. This does not mean that we should be subject to methodological individualism. Conventional economics so far has assumed that economic agents are completely rational in the sense that they are capable of solving any maximization problem. In reality, our ability of rational calculation is very limited, and it is now widely accepted that we should not neglect the fact that any decision is made by a person with a limited rationality. Behavior theory should be reorganized from a choice theoretic paradigm to a habitual behavior paradigm. Each individual action can be interpreted as an instance of a patterned behavior that is in turn chosen from a pool of repertoires. An evolutionary viewpoint is suggestive for the ecology of behaviors. The effectiveness of a patterned behavior can be roughly estimated as an average of performance of each pattern. Each individual has its pool of behavior patterns. Some of them may be invented by himself or herself, but most of them are learned from other individuals. A whole process comes out as a result of mutual interaction of individual behaviors, and individual behavior is conditioned by the whole process. In this sense, there is a kind of feedback loop between individual behaviors and the total process, which I call micro-macro-loop. With this understanding in mind, we should come back to the analysis of an economy as a whole. An economy is a huge network of individual actions. If we abandon the maximization principle, traditional mathematical analysis can hardly be applicable. A new analytical tool is required. One of the promising candidates is agent-based-modeling or multi-agent-modeling. In this modeling, each agent's behavior is rather simple but the varieties of behaviors and their interactions give rise to a dynamics of very complicated nature. The process can only be traced by computer calculation. An agent should be complex in such a way that it has an internal model of the outer world. Enforced learning and evolutionary selection can be incorporated by the standard method of genetic algorithms and other methods.
社会・経済システム学会の論文
2000-11-10

システム・アプローチに欠けるもの : 経済学における反省(<特集>システム論を問いなおす)

スポンサーリンク

概要

著者

関連論文

スポンサーリンク